Is Kantar actually tracking sales share, rather than ownership? That seems to be the only way these two sets of numbers reconcile, and is plausible from the wording of the article.
I love the fact that the article you linked, which is about how Google have gained 1.1% marketshare, and Apple gained 0.7% marketshare, is called "Apple's U.S. Smartphone Market Share Rises" with a massive picture of an iPhone.
From the article, Samsung grew 1.2% of all phones (including feature phones). So I'd say they almost certainly did grow faster than Android as a whole. I wouldn't be that surprised if Android minus Samsung contracted slightly in the quarter.
It's not about "fanbois" (whatever the hell that means), it's about the fact that articles about Apple attract orders-of-magnitude more clicks, and PR-reposting sites like Mashable worship at the alter of pageviews.
> It's not about "fanbois" (whatever the hell that means)
I suspect you know exactly what that means, but you're being intentionally dense in order to make a point, although you have ended up doing so in a very transparent fashion.
It's not so much a "bump" as a cycle. They release a phone every year, this was that quarter. Clearly they won't sell the same number of units next quarter. Their competitors ship far more (and more varied) products and don't really show that kind of behavior. So next quarter we'll be treated to "iOS is shrinking!" posts which are equally silly. And next fall the circle of life will repeat.
As you correctly point out, the real news in this data is that Blackberry sales dropped by 80%, which is so shocking that I wonder if a digit got slipped somewhere.
"so shocking that I wonder if a digit got slipped somewhere."
Sounds like Verizon flipped a certain digit at RIM. Possibly a good amount of August 2012 iPhone syndrome as well, "don't buy a Blackberry now, the new ones are coming in January"
This dance is now familiar. Apple does better than 50% in their release quarter and then slowly sinks to something like 15%-20% in the months prior to the next phone's release.
The interesting thing is that this time Apple launched the iPhone 5 to more or less the whole world in the first 3 months and change and so most people think they're moving to a semi-annual release schedule.
There's a tendency in the "tech" press to focus on a very narrow idea of "share". When they say "market share" they are almost always talking about new unit sales. This is because it's a lot cheaper to just take reports from companies that issue them (as press releases) and then write up a story from that... and these companies are analyzing only the most recent sales (for the most party.)
Actual market share is something quite different, as it includes all of the previously sold, but still operating devices.
The share relevant to developers is still another matter, as it must take into account the operating system the developer is targetting and how many devices still in use are using that operating system.
So, if you're a developer for high end mobile games, you might require Android Jelly Bean and iOS 6. Thus the question of where to direct your resources would best be answered by "What is the addressable market size for these two operating systems?"
I don't know the answer to that, but too often the torrent of shallow media news talking about "Share" that only talks about the current sales is taken as some sort of relevant long term value.
It's like saying Las Vegas is cold because you took the temperature one day in January. Worse you took that temperature in a freezer, rather than out in the open air. (Eg you're not even measuring the same thing the conclusion is about.)
The share of phones sold in the quarter is of most interest to investors. The share of installed base is more interesting to developers.
One major difference between the two that is often overlooked is that if someone upgrades their old phone and stops using the old one, it counts toward the former, but not toward the latter. Developers only really care about new users, and users who are switching platforms.
"The share of phones sold in the quarter is of most interest to investors."
Actually, not exactly. Let's say Apple sold 30% of all smartphones in a quarter. Now, from an investor standpoint that may seem pretty bad. However, Apple has a much higher profit margin, so what's actually relevant to investors and shareholders from a strict, short-term ROI basis is the share of the total revenue and the share of the total profit. As it turns out, Apple's share of the smartphone market's revenue is around 80%, so even if they dipped down to 30% of all smartphone handsets sold they would still have about half of the revenue.
I have some apps doing better on Android, and some that do better on iPhone. Overall, iPhone does much better, but Android is gaining. Sometimes there is a bit of luck involved, and sometimes, the competition in a certain niche is significantly different between the 2 stores.
Global growth probably doesn't matter as much if your app isn't hasn't been localized to all of those places. For example, growth of Android in a country where most people don't speak English isn't going to do much for an app written for an English audience.
If Android's growth is X times the iPhone's growth, and Android had a bigger "market share" a year ago (according to the article's chart), shouldn't iPhone's share be lower than Android's?
You say this as if the press is using the term in a misleading or special way, but in business the term "market share" has always meant percentage of volume for a given time period, not including product still being used. People might use a term like "usage share" for that, being something completely different.
Not generally, because few would define a market in terms of a single day since it's not very useful. Usually a period of one year is assumed but can commonly be a quarter or a month.
But technically, if you defined it that way, sure it had the biggest market share for that day.
Exacly. A much better way to collect this kind of data is too see the usage of mobile browsers. If you could get this data from de 100 most visited websites, and make a medium, maybe you could get a better 'market share' number.
Even more precisely would be to collect IMEI numbers from mobile operators. In the absence of that I've seen some guerilla work where people drive around 'fake' 3G cell sites which slurps IMEI's from phones that ping them. Totally illegal of course but if you read the article on how an RPi was turned into a cell site [1] you can get an idea of how its done. Since you're not actually taking calls you don't raise the suspicions of users.
This report is not measuring new unit sales. "Kantar’s panel data is generated by more than 260,000 consumer research interviews on a year basis in the U.S. alone"
At what point does the DOJ declare that they have a dominant share of the smartphone market, and that there should be increased oversight of anti-competitive behavior?
Is Kantar actually tracking sales share, rather than ownership? That seems to be the only way these two sets of numbers reconcile, and is plausible from the wording of the article.